array to image using the matplotlib

Question

I am trying to predict the depths from a single. the prediction is in .mat structure

{'__header__': b'MATLAB 5.0 MAT-file Platform: nt, Created on: Mon Jun 24 20:53:06 2019', '__version__': '1.0', '__globals__': [], 'mat': array

 ([[[[45.593273 ],
     [44.393333 ],
     [45.71833  ],
     ...,
     [20.923761 ],
     [22.743704 ],
     [25.33584  ]],

    [[44.40423  ],
     [43.71836  ],
     [45.744335 ],
     ...,
     [17.319963 ],
     [18.493673 ],
     [19.687391 ]],

    [[45.414154 ],
     [45.360767 ],
     [46.82794  ],
     ...,
     [17.41372  ],
     [18.701914 ],
     [19.105356 ]],

    ...,

    [[ 6.8809724],
     [ 6.5185905],
     [ 6.602861 ],
     ...,
     [ 6.6090174],
     [ 6.587162 ],
     [ 7.3165007]],

    [[ 7.0013733],
     [ 6.384803 ],
     [ 6.6147423],
     ...,
     [ 6.46376  ],
     [ 6.434711 ],
     [ 7.4840846]],

    [[ 7.9435554],
     [ 6.6214004],
     [ 6.654035 ],
     ...,
     [ 7.4519753],
     [ 7.712751 ],
     [ 9.646708 ]]]], dtype=float32)}

The final element mat is of the array type. I want to take the mat array and plot using matplotlib.

When I get the type and shape of the above I get this

The type of array is <class 'numpy.ndarray'>. The shape of the array is (1, 96, 320, 1).

I am not able to understand what does this means (1, 96, 320, 1) never seen such kind of shape with 4 parameters.

import scipy.io as sio 
import numpy as np
import matplotlib.pyplot as plt
mat_content = sio.loadmat('2011_09_26_drive_0002_sync_image_0000000023_image_02.mat')
#print(mat_content)
print(mat_content.keys())
array_content = mat_content['mat']
print(array_content)
print('The type of array is ',type(array_content)) 
print('The shape of array is ', array_content.shape)
#plt.plot(array_content) 
#plt.show()

when I run the above code I get this error

Traceback (most recent call last):
  File "2.py", line 11, in <module>
    plt.plot(array_content)
  File "C:\Users\snanr\Anaconda3\lib\site-packages\matplotlib\pyplot.py", line 2811, in plot
    is not None else {}), **kwargs)
  File "C:\Users\snanr\Anaconda3\lib\site-packages\matplotlib\__init__.py", line 1810, in inner
    return func(ax, *args, **kwargs)
  File "C:\Users\snanr\Anaconda3\lib\site-packages\matplotlib\axes\_axes.py", line 1611, in plot
    for line in self._get_lines(*args, **kwargs):
  File "C:\Users\snanr\Anaconda3\lib\site-packages\matplotlib\axes\_base.py", line 393, in _grab_next_args
    yield from self._plot_args(this, kwargs)
  File "C:\Users\snanr\Anaconda3\lib\site-packages\matplotlib\axes\_base.py", line 370, in _plot_args
    x, y = self._xy_from_xy(x, y)
  File "C:\Users\snanr\Anaconda3\lib\site-packages\matplotlib\axes\_base.py", line 234, in _xy_from_xy
    "shapes {} and {}".format(x.shape, y.shape))
ValueError: x and y can be no greater than 2-D, but have shapes (1,) and (1, 96, 320, 1)

I expect the output to be able to use the array and plot the array as an image using matplotlib.

arr = array_content[0,:,:,0] or arr = np.squeeze(array_content) should remove the extra size 1 dimensions. — hpaulj
– hpaulj, Commented Aug 8, 2019 at 15:43
plt.plot expects a 2-D array to plot for a line plot. Your array is 4-D. Usually if you want to plot an image you use plt.imshow. I'm still not sure if it will be able to plot a 4-D array though — Juan C
– Juan C, Commented Aug 8, 2019 at 15:44

Simon Notley · Accepted Answer · 2019-08-08 15:52:17Z

1

It looks like your input array is four dimensional for some reason, but two of the dimensions are only 1-deep. taking a wild guess that the dimensions with >1 depth are those of interest you could try

array_content = mat_content['mat'][0,:,:,0]

This should slice your 4D array down to a 2D one. Quick snippet to illustrate this below:

import numpy

x=numpy.zeros([1,96,320,1])
print(x.shape)
print(x[0,:,:,0].shape)

answered Aug 8, 2019 at 15:52

Simon Notley

2,1363 gold badges14 silver badges21 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

py_physicist · Accepted Answer · 2019-08-08 16:11:03Z

So I am not aware if the library you are using to display your image matters, but there is a much cleaner solution that does not require you to slice and manipulate your array. The code I am going to include uses a library called PIL. If you are running your code from an environment do a quick Pip install of Pillow (pip install Pillow would be the command in your prompt).

The code I have takes an image, prints off the matrix, and displays the image on a matplotlib graph. The only thing you need to change in the code is the filename you are using. Make sure that you load the file name in properly and you should be good to go.

import numpy as np
from PIL import Image
import matplotlib.pyplot as plt
image_object = Image.open("file_path.jpg")
matrix = np.asarray(image_object)
print(matrix)
PIL_object = Image.fromarray(matrix)
plt.imshow(PIL_object)
plt.show()

If this is the wrong approach I apologize! Hope it helps!

Collectives™ on Stack Overflow

array to image using the matplotlib

2 Answers 2

Comments

Comments

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

Comments

Related